Soybean Growth and Developmental Stage-Driven Crop Coefficient Simulation.

Soybean (Glycine max L.[Merr]) crop ET coefficients (Kc) for scheduling irrigations vary by growth-stage occurrences (e.g. early, middle, and late season Kc values). Scoring of soybean phenological stages (V, vegetative and R, reproductive) in a given field requires averaging 10 to 20 plants and must be performed twice per week to obtain the timeliness required in the delivery of irrigation water. In this study we 1) simulated soybean growth stages, 2) compared the simulated and actual stage occurrence in the field, and 3) related the cumulative soil water depletion (simulated) to actual depletion in the field. SoyWater – a web-based irrigation decision tool that incorporates the phenology component of the soybean crop model SoySim – was used to simulate soybean phenology. The soybean Kc coefficients of 0.1, 0.1, 1.1, 1.1, 0.20 and 0.10 were assigned to the stages VE (emergence), R0 (floral induction), R2 (full bloom), R6 (full seed), R7 (beginning maturity) and R8 (full maturity), respectively. The Kc value after occurrence of the R7 (physiological maturity) was set to zero.  Monitoring of actual growth stages was done in performed in research fields and in Nebraska soybean producer fields. Hourly soil water potential was recorded in 1-, 2-, and 3-foot soil depths using granular matrix soil water sensors. Comparison of simulated soybean V and R stages with the corresponding actual V and R stages revealed a root mean square error (RMSE) of 3.6 days in research fields, whereas in the larger producer fields, the RMSE was somewhat greater (5.5 for the V-stages and 5.7 for the R-stages). Simulated accumulative soil water depletion and actual soil water depletion measured with soil water sensors were in reasonable agreement, despite soil water sensor quality differences, variance in the simulated estimates of soil water depletion, and variance in the estimation of the potential evapotranspiration (PET). Recent measurements of soybean Kc values in Nebraska (not yet published) will be used to fine-tune the current Kc values used in SoyWater for soybean irrigation scheduling.